You are on page 1of 10

Human Antibodies 18 (2009) 1–10 1

DOI 10.3233/HAB-2009-0196
IOS Press

Three decades of human monoclonal


antibodies: Past, present and future
developments
Michael Steinitz∗
The Department of Pathology, The Hebrew University-Hadassah Medical School, Jerusalem, Israel

Abstract. The hybridoma technique has been shown to be a most reproducible method for producing rodent monoclonal antibodies
but poor results were obtained when it was used for generating human monoclonal antibodies. For immunotherapy, murine
monoclonal antibodies are inadequate, whereas human monoclonal antibodies are virtually indispensable. Cellular, chemical,
genetic and molecular methods to generate human monoclonal antibodies have been developed. Most often, the monoclonal
antibodies for therapy are selected after deliberate vaccination, according to their high affinity towards an arbitrarily-chosen
epitope of a pathogen or cellular antigen and therefore the selection is obviously skewed. A major hindrance of the production
of therapeutic human monoclonal antibodies is the lack of an appropriate strategy to define and select the antibodies that would
be effective in vivo. In contrast to antibodies induced by vaccination, there has been only a marginal interest in monoclonal
antibodies which reflect antibodies of the innate immunity. In the future, human monoclonal antibodies that resemble antibodies
that are ubiquitously present in sera of healthy individuals might serve as novel therapies in diseases such as Alzheimer’s disease,
where no other therapy exists.

Keywords: Chimeric antibodies, Epstein-Barr virus, immunotherapy, human, humanized antibodies, hybridoma, molecular
engineering, monoclonal antibody, phage display, transgenic mice

1. Introduction antibodies. Because of the enormous clinical potential


initially ascribed to monoclonal antibodies, there have
In 1975 K öhler and Milstein succeeded in produ- been continuous attempts to construct human therapeu-
cing monoclonal antibodies in vitro [1]. This dramatic tic monoclonal antibodies. Today, the pharmacological
achievement could not have been accomplished with- industries are intensely involved in these developments.
out the extensive scientific and technical progress made The current approaches to production of human mono-
earlier by many laboratories worldwide. Although the clonal antibodies are still inadequate, as attested to by
original hybridoma technique has proved to be extreme- the continuous attempts to improve the techniques. In
ly reproducible, new strategies were introduced to im- addition, the methods offered today to test the in vivo
prove the production of monoclonal antibodies in ge- efficacy and effectiveness of human monoclonal anti-
neral and of human monoclonal antibodies in particu- bodies are not satisfactory. Finally, the fact that today
lar. Alternative techniques have also been developed only very few completely-humanized monoclonal anti-
to create native and even non-native, newly composed bodies are used in the clinical setting indicates that the
field is in its initial phase.
This review presents the main technical develop-
∗ Corresponding author: Michael Steinitz, The Department ments, starting with the initial hybridoma method and
of Pathology, The Hebrew University-Hadassah Medical School,
Jerusalem, 91120, POB 12272, Israel. Tel.: +972 2 6758204; Fax: ending with the current approaches for engendering
+972 2 6784010; E-mail: michaels@md.huji.ac.il. therapeutic human monoclonal antibodies.

ISSN 1093-2607/09/$17.00  2009 – IOS Press and the authors. All rights reserved
2 M. Steinitz / Three decades of human monoclonal antibodies: Past, present and future developments

2. Köhler and Milstein’s original method (Fig. 1) culties. First, there was no appropriate drug-resistant
myeloma cell line for immortalization of human B cells.
According to the classical hybridoma method, mice Second, antigen-sensitized human splenocytes are not
are immunized with a mixture of antigens, their spleen as readily accessible as those of rodents. Third, in con-
cells are fused with an immunoglobulin non-secreting, trast to rodents, the availability of antigen-sensitized
drug-sensitive cell-line and cloned in microwell plates. lymphocytes is extremely limited.
Supernatants derived from the microwells are analyzed In a recent publication, Adekar et al. [2] presented
for specific antibodies and cells from positive wells are a modification of the hybridoma method which sought
further grown and cloned. Unlimited amounts of mo- to overcome some of the limitations in making hu-
noclonal antibodies can thus be produced in vitro. The man monoclonal antibodies. They introduced in vitro
method is based on three principles: a) each B lympho- antigen-specific sensitization of human B cells before
cyte produces only one antibody (i.e., an antibody with hybridization with an appropriate myeloma cell partner
specific H and L chain-derived variable regions), b) the and successfully produced IgG anti-botulinum neuro-
lymphocytes used for fusion are derived from donors toxin antibodies. It is not clear whether the method is
that were sensitized with specific immunogens, and reproducible for other antigens too.
c) B cells can be immortalized into immunoglobulin-
secreting, continuously growing in vitro cell lines.
Köller and Milstein’s ingenious hybridoma tech- 3. The Epstein-Barr virus method (Fig. 2)
nique has been widely used to fulfill both scientific and
practical/medical objectives. The incredible success of In the early seventies, Dr. George Klein of the
the method, which guaranteed the production of murine Karolinska Institutet, Stockholm, studied the role
antibodies with seemingly all possible specificities, has played by Epstein-Barr virus (EBV) in Burkitt’s lym-
caused some of the inherent limitations related to the phoma. At that time EBV was shown to efficiently
application of these antibodies in the clinic to be over- immortalize in vitro human immunoglobulin surface-
looked. First, the antibodies are produced by murine positive B cells into immortalized lymphoblastoid cell
cells and thus differ from proteins made by human cells. lines (LCL’s) [3], provided anti-EBV cytotoxic T cells
Second, the specificity of the monoclonal antibodies, were first removed or inactivated. The emerging cell
i.e., the molecule epitope recognized by the antibodies lines which express nine virus-induced proteins, pre-
and their biological function, resembles those of the serve the characteristics of the initially virus-infected
rodent immune system. Thus, murine monoclonal an- cell. Indeed, the LCL’s secrete and even express cell-
tibodies bind only to the molecular epitopes which the surface immunoglobulin. At the end of 1974, I was
mouse identifies as immunogenic. Moreover, the lym- a post doctoral student with George Klein who put
phocytes used to make the hybridomas are usually B forward the idea that immortalization with EBV of
cells from a mouse immunized by a selected molecule antigen-committed B cells would probably establish
and via a specific route. Hence, the monoclonal an- cell lines that secrete human antibodies.
tibodies may not resemble those protective antibodies Obviously, immortalization of peripheral blood lym-
which are induced under natural circumstances. Al- phocytes (PBL) of an immunized human donor creates
though generally ignored, these limitations are critical a polyclonal cell culture. Because of the extremely
when certain therapeutic requirements are assigned to low frequency of antigen-specific B cells among pe-
the antibodies. ripheral blood lymphocytes, there is hardly any chance
A major disadvantage of the murine monoclonal an- of detecting these cells in the emerging immortalized
tibodies when used for therapeutic purposes is that they cell culture. We, therefore, first enriched for specific
are recognized in the patients as allogeneic proteins. antigen-committed cells and then infected the culture
Indeed, the human anti-mouse antibody (HAMA) re- with EBV. Enrichment of the specific cells was car-
sponse following treatment with mouse monoclonal an- ried out by selection of cells that expressed on their
tibodies rapidly inactivates and eliminates the latter. cell-surface the corresponding antibody. Indeed, se-
The hybridoma method, which was so efficiently ex- lection of antigen-binding cells before the viral infec-
ploited in rodents, failed when adopted for the produc- tion enabled us for the first time to establish a variety
tion of human monoclonal antibodies. The failure and of human monoclonal antibody secreting cell lines [4].
inadequacy of the method to create human hybridomas This original method was applied to produce IgM, IgG
were due to some obvious and some unexpected diffi- and IgA human monoclonal antibodies against a vari-
M. Steinitz / Three decades of human monoclonal antibodies: Past, present and future developments 3

Vaccination with a mixture


of A and B antigens Myeloma cells
(HGPRT-negative)

Splenocytes
Fusion

Tissue culture in HAT medium

Anti-A positive Anti-B positive


(ELISA) (ELISA)

Cloning Cloning

Expansion of anti-A and anti-B clones

Fig. 1. Hybridoma method for making murine monoclonal antibodies.

ety of antigens. Greater stability and increased amount tion [8]. However, this approach has so far not proved
of secreted antibody were obtained by fusion of the successful.
EBV-immortalized antibody secreting cells with an ap- Recently we, together with Dr. R. Laskov from the
propriate heterohybridoma [5]. Recently, the EBV Hebrew University, Hadassah Medical School, found
method was further improved by using CpG to stim- that the occasional loss of antibody activity occurring
ulate the peripheral blood lymphocytes simultaneous- during prolonged in vitro tissue culture of LCL’s is due,
ly with infection with EBV [6]. The stimulation with at least partly, to induction of a specific enzyme, name-
CpG extends the target B-cell population which is EBV- ly activation-induced cytidine deaminase (AID) [9,10].
immortalizable so that it consists virtually of all the B This enzyme, which plays an important role in somat-
cells within the PBL, including memory CD27 posi- ic hypermutation (SHM) and isotype switching in B
cells [11], is upregulated in EBV infected cells. The
tive cells but excluding plasma cells. The frequency of
outcome of SHM in the in vivo maturation of B cells
antigen-committed B cells in PBL even in the blood of
is the selection of antibodies with a higher affinity,
an immunized individual is very low and therefore it
whereas in vitro such focused selection does not occur.
is crucial to immortalize as many B cells as possible.
Thus, mutations may induce a decrease or an increase
The efficiency of EBV-induced immortalization is sig- in antibody affinity during prolonged LCL culture.
nificantly heightened if the cells are infected with an
excess of virus [7].
Vaccination of healthy individuals to generate 4. Chemical and molecular methods: Chimeric
antigen-committed B cells for making monoclonal an- monoclonal antibodies
tibodies is not permitted but there is a plausible alterna-
tive strategy to prepare these cells by antigen-specific The facility with which murine monoclonal antibo-
sensitization of PBL in vitro prior to EBV immortaliza- dies can be produced using K öhler and Milstein’s
4 M. Steinitz / Three decades of human monoclonal antibodies: Past, present and future developments

Peripheral blood lymphocytes (PBL) 5. Humanized monoclonal antibodies

The molecular methods developed and improved in


Selection of antigen-committed B cells the past two decades and the greater comprehension
of the structure and function of the different antibody
domains led to novel revolutionary approaches to the
Infection with Epstein-Barr virus (EBV, B95-8) production of monoclonal antibodies. Whereas the hy-
bridoma and EBV methods facilitate immortalization
of specific antibody-committed B cells, the molecular
Tissue culture in microplates with CpG, Cyclosporine A techniques focus on immortalization of genes corre-
and sponding to specific antibodies. Molecular techniques
irradiated feeder cells (PBL) were used to further eliminate those portions in the
murine immunoglobulin chains that are not involved
in the binding of antigen and to replace them with the
corresponding human sequences. Complementarity-
determining regions (CDR’s) within the variable re-
gions of both the heavy and light chains play a promi-
nent role in the binding specificity of the antibody.
DNA fragments that correspond to the CDR’s were
Antibody-positive supernatants
(ELISA) grafted into the framework of human immunoglobulin
genes using molecular methods [13]. In addition, re-
placement of some amino acid residues in the constant
regions of the “recipient” human immunoglobulin with
Cloning the corresponding amino acids of the mouse “parental”
monoclonal antibody proved advantageous [14].
Thus, humanized antibodies retain the specificity and
Expansion of antibody secreting clones binding affinity of the “parental” murine monoclonal
antibody, they exhibit reduced immunogenicity in hu-
Fig. 2. Outline of Epstein-Barr virus method to engender human mans and they acquire biological functions of choice.
monoclonal antibodies.

method and the serious constraints these antibodies


6. Molecularly-engineered, completely human
impose upon their use in the clinical setting led to
monoclonal antibodies (Fig. 3)
new approaches. Methods were developed to convert,
at least partly, the readily available rodent monoclon-
al antibodies into antibodies with predominantly hu- Development of the PCR method made it possible to
man immunoglobulin chains, preserving those parts of amplify the entire immunoglobulin genes or some of
the murine antibody which correspond to the antigen- their components. Subsequently, these genes could be
binding sites. introduced into a variety of cells to produce the corre-
Initially, the Fc portion of the antibody molecule, sponding antibodies. Methods were applied to prepare
which dictates the functions of the antibody, was che- libraries of plasmids with the cDNA’s of heavy and light
mically exchanged with a human constant portion [12], chains from PBL’s and even from single cells derived
giving rise to chimeric monoclonal antibodies. The from naı̈ve and immunized human donors. The combi-
antigen specificity of the chimeric antibodies is iden- natorial libraries were used to transfect bacteria which,
tical to that of the initial mouse monoclonal antibody, in turn, were seeded on appropriate drug-supplemented
whereas the functions are determined solely by the hu- agar medium. Colonies producing active antibodies
man Fc domain. In comparison to the original mouse were thus detected and isolated. This method facili-
monoclonal antibodies, the chimeric molecules are less tates the formation even of antibodies which have never
“murine” and they therefore induce a significantly de- been made in vivo [15]. The main disadvantage of the
creased HAMA response in human recipients. How- method is that it does not provide appropriate ready-to-
ever, the remaining immunogenicity renders even these use tools for the selection and isolation of the desired
antibodies non-tolerable. high affinity specific antibodies.
M. Steinitz / Three decades of human monoclonal antibodies: Past, present and future developments 5

Peripheral blood lymphocytes (PBL s) / single cell brary is constructed from diverse variable regions of
from immune / naive donor the immunoglobulin genes (i.e., single chain of joined
heavy and light variable fragments (scFv), Fab frag-
ments or single VH or VL domains). Each library is
Isolation of immunoglobulin genes mRNA fashioned from cDNA derived from immune or naive
B cells. The DNA library is ligated into a surface
protein gene (gene III) of a bacteriophage. The bac-
Creation of cDNA and amplification by PCR
teriophages thus display on their surface the antibody
constructs fused with the surface protein. Phages ex-
pressing the required specificities are readily isolated
Ligation into H and L vectors
and enriched, using antigen-conjugated affinity bind-
ing columns. Eluted phages, similarly re-selected, are
Assembly of combined H and L chain libraries used to infect Escherichia coli to produce the mono-
clonal antibody construct. Alternately, the genes of
the specific antibody are excised and cloned into whole
Transfection of bacteria human IgG expression vectors and subsequently trans-
fected into appropriate cells to produce fully human
Seeding of bacteria on agar plates
monoclonal antibodies. Since this combinatory library
randomly matches the V regions of the heavy and light
chains, the resulting products include not only combi-
nations that were expressed by B cells in vivo but even
novel combinations (“non-native”) which never exist-
ed before. The libraries prepared from both na ı̈ve and
immunized human donors enable the selection of high
affinity antibodies. The antibody products are skewed
Colony screening for antigen binding as they depend on the specific antigen/epitope and en-
richment methods used to isolate the bacteriophages.
Fig. 3. Molecularly engineered monoclonal antibodies made by The phage display technique has the drawback that
combined H and L chain libraries.
it does not facilitate isolation of the genes correspond-
Jespers et al. [16] used molecular methods to pro- ing to the full-length IgG molecule. Mazor and co-
duce fully human monoclonal antibodies by pairing the workers [20] constructed a single plasmid which in
human VL chain with the VH of a mouse monoclonal bacteria can express both light and heavy full-length
antibody and then pairing the ”successful” human V L chains. The assembled antibody produced by the bac-
chain with a repertoire of human V H chains. It was teria is bound by an Fc-binding protein that is anchored
thus possible to come up with a human antibody with in the membrane of the bacteria. Thus, bacteria ex-
the specificity of the original murine antibody. pressing an antibody with a required specificity are
Molecular engineering of antibodies enables their readily isolated by affinity chromatography using solid
production in cells derived from a variety of sources, phase-conjugated antigen.
including bacteria [17] and plants [18]. The glyco-
sylation pattern of proteins, which is cell-dependent,
is crucial for therapeutic antibodies and therefore the 8. The transgenic mouse approach (Fig. 5)
type of cells in which they are generated is of major
importance. Köhler and Milstein’s classical hybridoma technique
in rodents has proved to be an extremely reproducible,
straightforward and problem-free method but has been
7. Phage display method (Fig. 4) rather disappointing in the human context. The fact
that production of human monoclonal antibodies by the
Phage display technology [19] is an ingenious ap- available techniques has been by far more complicated,
proach which provides the tools for creating and effi- inspired the development of a novel technique exploit-
ciently isolating high affinity recombinant native and ing transgenic mice [21]. Accordingly, the genes of
non-native monoclonal antibodies. A recombinant li- the heavy and light chains of human immunoglobulins
6 M. Steinitz / Three decades of human monoclonal antibodies: Past, present and future developments

Mixture of recombinant
Recombinant phages
phage mixture

Antigen -
Antigen-conjugated
conjugated matrix
matrix

Eluted
Elutedfraction:
:
fraction: Flow-through
- fraction:
:
Antigen binding phages
Antigen-binding phages Non-bound
- phages

Screening for desired antibody activity

Fig. 4. Outline of phage display method for producing monoclonal antibodies. The schematic (top) depicts the engineered bacteriophage (with
Enco’s permission).

replaced those of the mouse genes. Upon vaccination, mally present in human serum. Fourth, the durability
these knockout/knockin mice produced human antibod- of the foreign human chromosomal material is of ma-
ies and their spleens were used to make human mon- jor concern. A disturbing drawback is that biological
oclonal antibodies, applying the conventional hybrido- industries are the proprietors of the knockout/knockin
ma technique. Similarly, a specific TransChromo tech- mice, which are, therefore, not freely available to the
nology was developed whereby human chromosomes scientific community.
14 and 2 were introduced into heavy and light chain
gene-deficient knockout mice [22]. At the moment
some problems related to the technology remain unre- 9. Choice of therapeutic human monoclonal
solved. First, the human immunoglobulin genes trans- antibodies
ferred into the knockout mice are incomplete. Second,
the Ig-“humanized” transgenic mice possess murine T Since monoclonal antibodies resemble but a very
cells and therefore their humoral response is not purely small fraction of the antibody repertoire induced by any
human-specific. Third, in this system glycosylation of antigenic challenge, the criteria used for their selection
the human antibodies is mouse-specific. Thus, if the are of major significance. Needless to say, the genera-
antibodies are intended for immunotherapy they will tion and, even more so, the preferred choice of human
be recognized by anti-Galα 1-3Gal antibodies [23] nor- therapeutic monoclonal antibodies or the correspond-
M. Steinitz / Three decades of human monoclonal antibodies: Past, present and future developments 7

III
I Immunization with a mixture
Isolation of of A and B antigens
H and L
human genes /
chromosomes

IV
Fusion of splenocytes
with myeloma
cell line

II
Transgenic mouse with
human Ig genes / chromosomes

Anti-A positive Anti-B positive


(ELISA) (ELISA)

Fig. 5. Production of human monoclonal antibodies by transgenic mice.

ing genes, depend on specific considerations and crite- Innate immunity-related antibodies play a role also
ria which are only partly shared with those applied in in the removal of cellular waste, and modified and trans-
the context of murine monoclonal antibodies. Gener- formed cells. Vollmar and Brandlein [28] showed that
ally, murine monoclonal antibodies are evaluated pri- the natural anti-tumor antibodies in normal serum are
marily by the molecular epitope of their specific anti- germ-line-coded natural IgM antibodies. These IgMs
gen and by their affinity. These principles may be valid preferentially bind to carbohydrate epitopes on post-
for human therapeutic monoclonal antibodies, yet it is transcriptionally modified surface receptors and induce
their biological functions in vivo which are of major cancer-specific apoptosis.
importance. Indeed, selection of monoclonal antibod- B cells that produce innate immunity-related anti-
ies according to their specificity might be misleading. bodies might be a promising source for generating ther-
For example, in sera of HIV patients some antibod- apeutic monoclonal antibodies.
ies against HIV antagonize other anti-virus protective The PBL of individuals who are deliberately vacci-
antibodies [24].
nated against specific pathogens are used as a source of
Natural antibodies against a variety of pathogens
B cells to make anti-pathogen protective monoclonal
and altered cells are present in the sera of normal hu-
antibodies. However, since there is no comprehensive
mans. The antibodies against bacteria and viruses are
analysis on a single cell level, the choice of the “appro-
either natural (i.e., innate immunity) or induced after
an earlier encounter with the specific pathogen (i.e., priate” B cells for immortalization is uncertain. Wram-
acquired immunity). Protective anti-pathogen or anti- mert et al. [29] showed that vaccination against in-
altered cell antibodies have been tailored for millions of fluenza induces a transient IgG anti-influenza response
years to serve as a safeguard against common pathogens which has a distinct clonality and kinetics. They al-
and transformed cells. The specificity and the affini- so found that in influenza-vaccinated humans the re-
ty of these antibodies are not necessarily identical to sponse of memory B cells and antibody-secreting plas-
the high-affinity and epitope-selected monoclonal an- ma cells (ASC) is organ-specific. Wrammer and co-
tibodies which are made so efficiently in vitro. This workers who successfully prepared and analyzed re-
holds true for a variety of antibodies, as for example combinant anti-influenza fully human antibodies de-
anti-influenza antibodies [25], anti-HCV protective an- rived from single ASC’s, showed that the anti-influenza
tibodies [26] and HIV-neutralizing antibodies [27]. vaccine-induced response has a limited antibody reper-
8 M. Steinitz / Three decades of human monoclonal antibodies: Past, present and future developments

toire [29]. The apparent restricted response, as reflect- ies are entirely of a human source. The monoclonal an-
ed in the emerging monoclonal antibodies [29], is prob- tibodies used today in the clinic are prepared from vac-
ably skewed and differs from that of the authentic in vi- cinated donors and reflect an acquired humoral immune
vo response. Indeed, in contrast to these results, a much response. Today, there is only a vague understanding of
broader repertoire of human anti-influenza antibodies the role played by “natural” antibodies in the homeosta-
was obtained using the Symplex technology [30]. This sis of healthy individuals. The majority of the “natural”
technology, which applies both cellular and molecu- antibodies in the serum are induced by unknown and
lar procedures, allows identification of high-affinity an- non-defined antigenic challenges. Even if the molecu-
tibodies from immunized or naturally-immunized in- lar targets of the antibodies are known, their function
dividuals and enables rapid direct cloning of the im- is obscure. It is anticipated that some of the antibodies
munoglobulin genes from single antibody-producing play a role in the homeostasis of normal molecules and
cells. their breakdown products, in removal of waste and aged
In a recent elaborate multistep study, Kurosawa et al. normal and transformed cells. Obviously, the source
tried to outline a method for the selection of anti-tumor of the cells to be used to make monoclonal antibodies
protective human monoclonal antibodies [31]. First, a resembling “natural”, possibly protective antibodies, is
phage antibody display library was constructed from a B lymphocytes from non-vaccinated healthy individ-
healthy individual to produce a repertoire of recombi- uals. The frequency of these B cells in normal indi-
nant antibodies. Antibodies binding to a variety of tu- viduals can be readily deduced from the titre of the
mor cell lines were then identified. Second, further se- corresponding antibodies in intravenous immunoglob-
lection allowed the isolation of antibodies that stained a ulin (IVIG) preparations, the commercially-available
variety of malignant cells. Finally, functional assays in pooled human IgG. In the future, human monoclonal
vitro (anti-tumor ADCC assay) and in vivo (anti-tumor antibodies resembling widespread “natural” antibodies
reaction in athymic mice) were used to choose anti- might offer promising novel reagents for immunother-
bodies of possible therapeutic potential. This laborious apy.
procedure which integrated molecular technology, cell Novel developments in clinical research point to
selection and functional assays, outlines a comprehen- additional areas where human monoclonal antibodies
sive straightforward procedure for isolation of human might be applied as passive vaccination, such as dis-
therapeutic antibodies. eases of the central nervous system (i.e., Alzheimer’s
Human monoclonal antibodies that resemble in vivo disease (AD)) and autoimmune diseases (i.e., systemic
genuine antibodies are of importance when considered lupus erythematosus). In a murine model of AD it has
as candidates for therapeutic antibodies and they are in- been shown that passive and active immunotherapy a)
dispensable for basic research on autoimmune diseases. eliminates the typical amyloid beta (Aβ) plaques which
The in vitro produced RF-AN monoclonal antibody characterize also the brain of AD patients and b) im-
(autoimmune IgM anti-IgG antibody, i.e., rheumatoid proves the cognitive behavior of the sick mice [33].
factor) was established in my laboratory by an EBV- The presence of anti-Aβ IgG antibodies in the serum
immortalized B cell derived from a rheumatoid arthri- of all individuals [34] is an indication that these anti-
tis patient [32]. This antibody, which was the first au- bodies play an important role in the homeostasis of the
toimmune antibody produced in vitro, has been studied amyloid precursor protein (APP). This assumption is
intensively. The sole authentic source for pathogenic backed by the finding that there is an improvement in
autoimmune antibodies would be B cells derived di- the cognitive behavior of AD patients who are treated
rectly from non-treated patients. In contrast, autoim- with IVIG [35,36].
mune monoclonal antibodies produced using molecu- We established human monoclonal anti-Aβ antibody
lar techniques do not necessarily resemble the authentic producing LCL’s by EBV-infection of B cells derived
pathogenic antibodies. from healthy blood donors [37]. These antibodies rep-
resent genuine antibodies that presumably participate
in the homeostasis of certain form(s) of the amyloid and
10. Future developments APP in healthy individuals. The anti-Aβ monoclonal
antibodies bind to the N-terminal of the 43 amino acid-
Human monoclonal antibodies have entered the clin- long Aβ amyloid. In vitro they bind to Aβ and not to
ic primarily as a passive vaccination in malignancies any other type of amyloid and they specifically stain
and inflammatory diseases. Only few of these antibod- amyloid plaques in brain sections derived from AD pa-
M. Steinitz / Three decades of human monoclonal antibodies: Past, present and future developments 9

tients. If the clinical trials will convincingly confirm [2] S.P. Adekar, R.M. Jones, M.D. Elias, F.H. Al-Saleem, M.J.
the encouraging results of the murine model of AD, hu- Root, L.L. Simpson and S.K. Dessain, Hybridoma popula-
tions enriched for affinity-matured human IgGs yield high-
man therapeutic monoclonal antibodies derived from affinity antibodies specific for botulinum neurotoxins, J Im-
healthy individuals may prove to be promising reagents munol Methods 333 (2008), 156–166.
for future therapy. [3] K. Nilsson, G. Klein, W. Henle and G. Henle, The estab-
Lupus is an autoimmune disease characterized by the lishment of lymphoblastoid lines from adult and fetal human
lymphoid tissue and its dependence on EBV, Int J Cancer 8
presence of a variety of autoimmune antibodies. Some (1971), 443–450.
of the patients develop a psychiatric syndrome that is [4] M. Steinitz, G. Klein, S. Koskimies and O. Mäkelä, EBV virus
associated with a high serum titre of anti-ribosomal induced B lymphocyte cell lines producing specific antibody,
proteins (i.e.,anti-P antibodies). Recently Matus et Nature 269 (1977), 420–422.
[5] D. Kozbor, Human anti-tetanus toxoid monoclonal antibody
al. [38] showed that anti-P antibodies recognize an in- secreted by EBV-transformed human B cells fused with murine
tegral membrane molecule present on neurons. This myeloma, Hybridoma 1 (1982), 323–328.
protein is preferentially distributed in the brain in areas [6] E. Traggiai, S. Becker, K. Subbarao, L. Kolesnikova, Y. Ue-
involved in memory, cognition, and emotion. In tissue matsu, M.R. Gismondo, B.R. Murphy and R. Rappuoli, An
efficient method to make human monoclonal antibodies from
culture, the anti-P antibodies induce apoptosis of neu- memory B cells: potent neutralization of SARS coronavirus,
ral cells whereas astrocytes, which do not express the Nat Med 10 (2004), 871–875.
protein, are not affected by the antibodies. The find- [7] C. Shannon-Lowe, G. Baldwin, R. Feederle, A. Bell, A. Rick-
ing that IVIG has a high titre of anti-P IgG antibod- inson and H.J. Delecluse, Epstein-Barr virus-induced B-cell
transformation: quantitating events from virus binding to cell
ies (Steinitz, unpublished results) points to a possible outgrowth, J Gen Virol 86 (2005), 3009–3019.
inhibitory or blocking role which they might play in [8] M. Ohlin, A.P. Broliden, L. Danielsson, B. Wahren, J. Rosen,
the homeostasis of healthy individuals. If certain anti- M. Jondal and C.A. Borrebaeck., Human monoclonal antibod-
P antibodies can indeed antagonize the effect of the ies against a recombinant HIV envelope antigen produced by
primary in vitro immunization. Characterization and epitope
pathogenic anti-P antibodies in lupus patients, produc- mapping, Immunology 68 (1989), 325–331.
tion of the former might provide an attractive reagent [9] I. Chezar, L. Lobel-Lavi, M. Steinitz and R. Laskov, Ongo-
for novel immunotherapy. Towards this end we have ing somatic hypermutation of the rearranged VH but not of
the V-lambda gene in EBV- transformed rheumatoid factor-
applied the EBV method to produce anti-P antibody
producing lymphoblastoid cell line, Mol Immunol 46 (2008),
secreting LCL’s. 80–90.
The humoral immune system in higher vertebrates [10] Y. Gil, S. Levy-Nabot, M. Steinitz and R. Laskov, Somatic
is unique in its ability to generate highly diverse anti- mutations and activation-induced cytidine deaminase (AID)
expression in established rheumatoid factor-producing lym-
body responses against pathogens, as well as against
phoblastoid cell line, Mol Immunol 44 (2007), 494–505.
certain malignancies. The likelihood of discovering a [11] J.U. Peled, F.L. Kuang, M.D. Iglesias-Ussel, S. Roa, S.L.
single “appropriate” monoclonal antibody is probably Kalis, M.F. Goodman and M.D. Scharff, The biochemistry of
far fetched. In nature, an efficient immune response somatic hypermutation, Annu Rev Immunol (2008), 26 481–
511.
against pathogens is never monoclonal, probably be- [12] G.L. Boulianne, N. Hozumi and M.J. Shulman, Production
cause the polyclonal response is by far more effective. of functional chimaeric mouse/human antibody, Nature 312
Thus, we may expect that in the future immunotherapy (1984), 643–646.
will involve a mixture of selected monoclonal antibo- [13] P.T. Jones, H.P. Dear, J. Foote, M.S. Neuberger and G. Win-
ter, Replacing the complementarity-determining regions in a
dies with a variety of specificities. human antibody with those from a mouse, Nature 321 (1986),
522–525.
[14] P.J. Carter, Potent antibody therapeutics by design, Nat Rev
Acknowledgements Immunol 6 (2006), 343–357.
[15] L. Sastry, Alting-Mees, W.D. Huse, J.M. Short, J.A. Sorge,
B.N .Hay, K.D. Janda, S.J. Benkovic and R.A. Lerner, Cloning
This study was supported by the Ernst and Lili of the immunological repertoire in Escherichia coli for gen-
Rosenhain Fund, Bruxelles, Belgium. We thank Dr. eration of monoclonal catalytic antibodies: construction of a
Alexandra Mahler for her editorial assistance. heavy chain variable region-specific cDNA library, Proc Natl
Acad Sci USA 86 (1989), 5728–5732.
[16] L.S. Jespers, A. Roberts, S.M. Mahler, G. Winter and H.R.
Hoogenboom, Guiding the selection of human antibodies from
References phage display repertoires to a single epitope of an antigen,
Biotechnology (NY) 12 (1994), 899–903.
[1] G. Kohler and C. Milstein, Continuous cultures of fused [17] W.D. Huse, L. Sastry, S.A. Iverson, A.S. Kang, M. Alting-
cells secreting antibody of predefined specificity, Nature 256 Mees, D.R. Burton, S.J. Benkovic and R.A. Lerner, Genera-
(1975), 495–497.
10 M. Steinitz / Three decades of human monoclonal antibodies: Past, present and future developments

tion of a large combinatorial library of the immunoglobulin C. Larsen, N.Y. Zheng, I. Mays, L. Garman, C. Helms, J.
repertoire in phage lambda, Science 246 (1989), 1275–1281. James, G.M. Air, J.D. Capra, R. Ahmed and P.C. Wilson,
[18] M. Pujol, N.I. Ramirez, M. Ayala, J.V. Gavilondo, R. Valdes, Rapid cloning of high-affinity human monoclonal antibodies
M. Rodriguez, J. Brito, S. Padilla, L. Gomez, B. Reyes, R. against influenza virus, Nature 453 (2008), 667–671.
Peral, M. Perez, J.L. Marcelo, L. Mila, F.R.Sanchez, R. Paez, [30] P.J. Meijer, P.S. Andersen, M. Haahr Hansen, L. Steinaa,
J.A. Cremata, G. Enriquez, O. Mendoza, M. Ortega and C. A. Jensen, J. Lantto, M.B. Oleksiewicz, K. Tengbjerg, T.R.
Borroto, An integral approach towards a practical application Poulsen, V.W. Coljee, S. Bregenholt, J.S. Haurum and L.S.
for a plant-made monoclonal antibody in vaccine purification, Nielsen, Isolation of human antibody repertoires with preser-
Vaccine 23 (2005), 1833–1837. vation of the natural heavy and light chain pairing, J Mol Bio
[19] J.D. Marks, H.R. Hoogenboom, T.P. Bonnert, J. McCafferty, 358 (2006), 764–772.
A.D. Griffiths and G.Winter, By-passing immunization. Hu- [31] G. Kurosawa, Y. Akahori, M. Morita, M. Sumitomo, N. Sato,
man antibodies from V-gene libraries displayed on phage, J C. Muramatsu, K. Eguchi, K. Matsuda, A. Takasaki, M. Tana-
Mol Biol 222 (1991), 581–597. ka, Y. Iba, S. Hamada-Tsutsumi, Y. Ukai, M. Shiraishi, K.
[20] Y. Mazor, T. Van Blarcom, R. Mabry, B. Iverson and G. Suzuki, M. Kurosawa, S. Fujiyama, N. Takahashi, R. Kato,
Georgiou, Isolation of engineered, full-length antibodies from Y. Mizoguchi, M. Shamoto, H. Tsuda, M. Sugiura, Y. Hattori,
libraries expressed in Escherichia coli, Nat Biotechnol 25 S. Miyakawa, R. Shiroki, K. Hoshinaga, N. Hayashi, A. Sug-
(2007), 563–565. ioka and Y. Kurosawa, Comprehensive screening for antigens
[21] A. Jakobovits, G.J. Vergara, J.L. Kennedy, J.F. Hales, R.P. overexpressed on carcinomas via isolation of human mAbs
McGuinness, D.E. Casentini-Borocz, D.G. Brenner and G.R. that may be therapeutic, Proc Natl Acad Sci USA 105 (2008),
Otten, Analysis of homozygous mutant chimeric mice: dele- 7287–7292.
tion of the immunoglobulin heavy-chain joining region blocks [32] M. Steinitz, G. Izak, S. Cohen, M. Ehrenfeld and I. Flechner,
B-cell development and antibody production, Proc Natl Acad Continuous production of monoclonal rheumatoid factor by
Sci USA 90 (1993), 2551–2555. EBV-transformed lymphocytes, Nature 287 (1980), 443–445.
[22] I. Ishida, K. Tomizuka, H. Yoshida and Y. Kuroiwa, Tran- [33] D. Schenk, R. Barbour, W. Dunn, G. Gordon, H. Grajeda,
sChromo Mouse, Biotechnol Genet Eng Rev (2002), 73–82. T. Guido, K. Hu, J. Huang, K. Johnson-Wood, K. Khan, D.
[23] C.K. Borrebaeck, A.C. Malmborg and M. Ohlin, Does en- Kholodenko, M. Lee, Z. LiaoI, Lieberburg, R. Motter, L. Mut-
dogenous glycosylation prevent the use of mouse monoclonal ter, F. Soriano, G. Shopp, N. Vasquez, C. Vandevert, S. Walk-
antibodies as cancer therapeutics? Immunol Today 14 (1993), er, M. Wogulis, T. Yednock, D. Games and P. Seubert, Immu-
477–479. nization with amyloid-beta attenuates Alzheimer-disease-like
[24] C.P. Krachmarov, W.J. Honnen, S.C. Kayman, M.K. Gorny, pathology in the PDAPP mouse, Nature 400 (1999), 173–177.
S. Zolla-Pazner and A. Pinter, Factors determining the breadth [34] P. Szabo, N. Relkin and M.E. Weksler, Natural human an-
and potency of neutralization by V3- specific human mono- tibodies to amyloid beta peptide, Autoimmun Rev 7 (2008),
clonal antibodies derived from subjects infected with clade A 415–420.
or clade B strains of human immunodeficiency virus type 1, J [35] N.R. Relkin, P. Szabo, B. Adamiak, T. Burgut, C. Monthe,
Viro 80 (2006), 7127–7135. R.W. Lent, S. Younkin, L. Younkin, R. Schiff and M.E. Wek-
[25] N. Baumgarth, J.W. Tung and L.A. Herzenberg, Inherent sler, 18-Month study of intravenous immunoglobulin for treat-
specificities in natural antibodies: a key to immune defense ment of mild Alzheimer disease, Neurobiol Aging (2008).
against pathogen invasion, Springer Semin Immunopathol 26 [36] G. Istrin, E. Bosis and B. Solomon, Intravenous immunoglob-
(2005), 347–362. ulin enhances the clearance of fibrillar amyloid-beta peptide,
[26] R. Burioni, M. Perotti, N. Mancini and M. Clementi, Per- J Neurosci Res 84 (2006), 434–443.
spectives for the utilization of neutralizing human monoclonal [37] V. Geylis, V. Kourilov, Z. Meiner,I. Nennesmo, N. Bog-
antibodies as anti-HCV drugs, J Hepatol 49 (2008), 299–300. danovic and M. Steinitz, Human monoclonal antibodies
[27] R.L. Stanfield, M.K. Gorny, S. Zolla-Pazner and I.A. Wilson, against amyloid-beta from healthy adults, Neurobiol Aging 26
Crystal structures of human immunodeficiency virus type 1 (2005), 597–606.
(HIV-1) neutralizing antibody 2219 in complex with three [38] S. Matus, P.V. Burgos, M. Bravo-Zehnder, R. Kraft, O.H.
different V3 peptides reveal a new binding mode for HIV-1 Porras, P. Farias, L.F. Barros, F. Torrealba, L. Massardo, S.
cross-reactivity, J Virol 80 (2006), 6093–6105. Jacobelli and A. Gonzalez, Antiribosomal-P autoantibodies
[28] H.P. Vollmers and S. Brandlein, Natural antibodies and cancer, from psychiatric lupus target a novel neuronal surface protein
J Autoimmun 29 (2007), 295–302. causing calcium influx and apoptosis, J Exp Med 204 (2007),
[29] J. Wrammert, K. Smith, J. Miller, W.A. Langley, K. Kokko, 3221–3234.

You might also like